Ilya Sutskever is a computer scientist widely regarded as one of the most influential researchers of the modern Deep learning era. As a graduate student under Geoffrey Hinton at the University of Toronto, he co-created AlexNet with Alex Krizhevsky, the convolutional neural network whose decisive win at the 2012 ImageNet competition is widely credited with starting the modern deep learning wave. He later co-authored the Seq2seq paper in 2014 with Oriol Vinyals and Quoc Le, introducing the sequence-to-sequence encoder-decoder framework that became foundational for machine translation and later influenced the design thinking behind the Transformer (architecture) architecture.
Co-founding and leading research at OpenAI
Sutskever was a co-founder of OpenAI in 2015 and served as its chief scientist for most of the following decade, playing a central technical role in the development of the GPT-2, GPT-3, and GPT-4 models and the research direction that led to ChatGPT. He was an early and consistent advocate inside OpenAI for the scaling hypothesis, that substantially larger models and datasets would yield qualitatively new capabilities, a bet that shaped much of OpenAI's research strategy through the GPT-3 and GPT-4 era.
The 2023 board crisis and departure
Sutskever was one of the OpenAI board members who voted in November 2023 to remove CEO Sam Altman, contributing to the OpenAI board crisis. He later stated publicly that he regretted his role in the decision and supported Altman's reinstatement. Sutskever remained at OpenAI for several months afterward but left the company in May 2024, alongside a broader wave of departures from OpenAI's safety-focused research staff that also included fellow researcher Jan Leike.
Safe Superintelligence Inc.
In June 2024, Sutskever co-founded Safe Superintelligence Inc. (SSI) with Daniel Gross and Daniel Levy, structured around a single, narrowly defined mission: developing Superintelligence safely, with no near-term commercial products and, according to Sutskever, no pressure to ship until the technology is ready. The company drew a multi-billion-dollar valuation on the strength of Sutskever's reputation and research record despite the absence of a conventional product roadmap, an unusual structure that reflected investor confidence in his standing as one of the field's most credited technical minds, alongside continued public focus on AI alignment and safety as the central challenge of advanced AI development.